DESCRIPTION: The concept of "environmental justice" plays an increasingly important role in developing environmental regulations to protect the public's health. This form of justice relates to equity among all segments of the population with respect to exposure to environmental hazard and to health outcomes. Subpopulations of particular interest include pregnant women, children, ethnic, and racial minorities, and other historically disadvantaged groups. While data availability, storage, and processing flexibility have increased enormously in recent years, development of appropriate statistical methodology has not kept pace. Exposure and health outcomes data can be displayed in striking maps via Geographic Information Systems (GIS), but the best way to develop numerical inputs for these displays is not yet clear. A four-year research program is proposed to develop and evaluate the following: (1) methods that exploit spatial structure to stabilize rate estimates based on small numbers of people and thereby to maintain geographic and demographic resolution of stabilized rates; (2) statistical methods and models to detect areas of elevated exposure to environmental hazards, areas of increased disease incidence, and areas of increased risk while controlling for known confounding factors in both time and space; (3) guidelines and methods to allow for the combining of misaligned data from several sources (e.g./ exposure data, outcome data, and demographic data) into a cogent collection for analysis; and (4) a suite of software routines implementing the developed methods that will serve as the prototype system of analysis for statisticians, epidemiologists and geographers investigating environmental justice with GIS.